Joint 3D Layout and Depth Prediction from a Single Indoor Panorama Image

Monte Carlo Scene Search for 3D Scene Understanding

“…Many methods working from panoramic images [50,60,62] and point clouds [20,33,43] rely on such priors. Methods which utilize supervised learning [57,31,62,59,30,36,50,60,62,54,55] depend on large-scale datasets, the creation of which is a challenge on its own. When performing layout estimation from point clouds as input data [43,6,20,33,32], one has to deal with incomplete and noisy scans as can be found in the ScanNet dataset [14].…”

“…3D scene understanding is a fundamental problem in Computer Vision [41,53]. In the case of indoor scenes, one usually aims at recognizing the objects and their properties such as their 3D pose and geometry [2,3,15], or the room layouts [57,31,62,59,30,36,50,60,62,54,55], or both [4,18,35,45,51,56]. With the development of deep learning approaches, the field has made a remarkable progress.…”

360° Optical Flow using Tangent Images

“…Panoramic Image Processing. Panoramic (aka 360°or spherical) images have been used in a wide variety of application areas, including depth estimation [19,20,24,34,41,42,44,50,51], room layout estimation [12,14,21,34,40,43,47], semantic segmentation [25,34,46,48], novel-view synthesis [5,16,28,45] and so on. The key problem when working with spherical images is to project them onto a regular 2D pixel grid for easy processing, as any projection introduces some kind of distortion -similar to maps of the world.…”

Pano3D: A Holistic Benchmark and a Solid Baseline for $360^o$ Depth Estimation

“…The recent availability of 360 o depth datasets out of stitched raw sensor data [1,7], 3D reconstruction renderings [56,55], and photorealistic synthetic scenes [34,53] has stimulated research in monocular 360 o depth estimation [44,13,45,26,52,43]. Still, the progress in monocular depth estimation has been mainly driven by research for * Indicates equal contribution.…”